Advancement on Breast Cancer Detection Using Medio-Lateral-Oblique (Mlo) and Cranio-Caudal (CC) Features

Abstract

The purpose of this paper is to diagnose accurately the breast cancer for which a a computer aided diagnostic system (CAD) is being proposed. In this paper two types of views are used to enhance diagnostic efficiency, such as cranio-caudal (CC) and medio-lateral-oblique (MLO). This paper involves segmentation, feature extraction and classification of images. Adaptive K means clustering method is used in segmentation to segment the two views from a mammogram image. The combination of conventional k-means clustering method and Gabor filter is employed in the feature extraction stage to extract the features of CC and MLO views. Finally, Knn classifier is used to classify the mammogram image into four ways, such as CC-Normal, CC-Malignant, MLO-Normal and MLO-malignant.